0
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-18 01:35:12 +00:00
yt-dlp/yt_dlp/extractor/aol.py
pukkandan 86e5f3ed2e
[cleanup] Upgrade syntax
Using https://github.com/asottile/pyupgrade

1. `__future__` imports and `coding: utf-8` were removed
2. Files were rewritten with `pyupgrade --py36-plus --keep-percent-format`
3. f-strings were cherry-picked from `pyupgrade --py36-plus`

Extractors are left untouched (except removing header) to avoid unnecessary merge conflicts
2022-04-12 05:32:51 +05:30

133 lines
5.3 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

import re
from .yahoo import YahooIE
from ..utils import (
ExtractorError,
int_or_none,
parse_qs,
url_or_none,
)
class AolIE(YahooIE):
IE_NAME = 'aol.com'
_VALID_URL = r'(?:aol-video:|https?://(?:www\.)?aol\.(?:com|ca|co\.uk|de|jp)/video/(?:[^/]+/)*)(?P<id>\d{9}|[0-9a-f]{24}|[0-9a-f]{8}-(?:[0-9a-f]{4}-){3}[0-9a-f]{12})'
_TESTS = [{
# video with 5min ID
'url': 'https://www.aol.com/video/view/u-s--official-warns-of-largest-ever-irs-phone-scam/518167793/',
'md5': '18ef68f48740e86ae94b98da815eec42',
'info_dict': {
'id': '518167793',
'ext': 'mp4',
'title': 'U.S. Official Warns Of \'Largest Ever\' IRS Phone Scam',
'description': 'A major phone scam has cost thousands of taxpayers more than $1 million, with less than a month until income tax returns are due to the IRS.',
'timestamp': 1395405060,
'upload_date': '20140321',
'uploader': 'Newsy Studio',
},
'params': {
# m3u8 download
'skip_download': True,
}
}, {
# video with vidible ID
'url': 'https://www.aol.com/video/view/netflix-is-raising-rates/5707d6b8e4b090497b04f706/',
'info_dict': {
'id': '5707d6b8e4b090497b04f706',
'ext': 'mp4',
'title': 'Netflix is Raising Rates',
'description': 'Netflix is rewarding millions of its long-standing members with an increase in cost. Veuers Carly Figueroa has more.',
'upload_date': '20160408',
'timestamp': 1460123280,
'uploader': 'Veuer',
},
'params': {
# m3u8 download
'skip_download': True,
}
}, {
'url': 'https://www.aol.com/video/view/park-bench-season-2-trailer/559a1b9be4b0c3bfad3357a7/',
'only_matching': True,
}, {
'url': 'https://www.aol.com/video/view/donald-trump-spokeswoman-tones-down-megyn-kelly-attacks/519442220/',
'only_matching': True,
}, {
'url': 'aol-video:5707d6b8e4b090497b04f706',
'only_matching': True,
}, {
'url': 'https://www.aol.com/video/playlist/PL8245/5ca79d19d21f1a04035db606/',
'only_matching': True,
}, {
'url': 'https://www.aol.ca/video/view/u-s-woman-s-family-arrested-for-murder-first-pinned-on-panhandler-police/5c7ccf45bc03931fa04b2fe1/',
'only_matching': True,
}, {
'url': 'https://www.aol.co.uk/video/view/-one-dead-and-22-hurt-in-bus-crash-/5cb3a6f3d21f1a072b457347/',
'only_matching': True,
}, {
'url': 'https://www.aol.de/video/view/eva-braun-privataufnahmen-von-hitlers-geliebter-werden-digitalisiert/5cb2d49de98ab54c113d3d5d/',
'only_matching': True,
}, {
'url': 'https://www.aol.jp/video/playlist/5a28e936a1334d000137da0c/5a28f3151e642219fde19831/',
'only_matching': True,
}, {
# Yahoo video
'url': 'https://www.aol.com/video/play/991e6700-ac02-11ea-99ff-357400036f61/24bbc846-3e30-3c46-915e-fe8ccd7fcc46/',
'only_matching': True,
}]
def _real_extract(self, url):
video_id = self._match_id(url)
if '-' in video_id:
return self._extract_yahoo_video(video_id, 'us')
response = self._download_json(
'https://feedapi.b2c.on.aol.com/v1.0/app/videos/aolon/%s/details' % video_id,
video_id)['response']
if response['statusText'] != 'Ok':
raise ExtractorError('%s said: %s' % (self.IE_NAME, response['statusText']), expected=True)
video_data = response['data']
formats = []
m3u8_url = url_or_none(video_data.get('videoMasterPlaylist'))
if m3u8_url:
formats.extend(self._extract_m3u8_formats(
m3u8_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
for rendition in video_data.get('renditions', []):
video_url = url_or_none(rendition.get('url'))
if not video_url:
continue
ext = rendition.get('format')
if ext == 'm3u8':
formats.extend(self._extract_m3u8_formats(
video_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
else:
f = {
'url': video_url,
'format_id': rendition.get('quality'),
}
mobj = re.search(r'(\d+)x(\d+)', video_url)
if mobj:
f.update({
'width': int(mobj.group(1)),
'height': int(mobj.group(2)),
})
else:
qs = parse_qs(video_url)
f.update({
'width': int_or_none(qs.get('w', [None])[0]),
'height': int_or_none(qs.get('h', [None])[0]),
})
formats.append(f)
self._sort_formats(formats)
return {
'id': video_id,
'title': video_data['title'],
'duration': int_or_none(video_data.get('duration')),
'timestamp': int_or_none(video_data.get('publishDate')),
'view_count': int_or_none(video_data.get('views')),
'description': video_data.get('description'),
'uploader': video_data.get('videoOwner'),
'formats': formats,
}